P935Can we predict the occurrence of persistent atrial fibrillation in large animal models to improve study efficiency?
N Denham,C M Pearman,G W P Madders,C E R Smith,A W Trafford,K M Dibb
DOI: https://doi.org/10.1093/europace/euaa162.077
2020-06-01
EP Europace
Abstract:Abstract Funding Acknowledgements British Heart Foundation Project Grant FS/17/54/33126 Background A key consideration when using animals in research is maximising experimental efficiency to minimise the number of animals required. Large animal models have proven an invaluable tool for establishing pathophysiological mechanisms underpinning atrial fibrillation (AF) and testing novel therapeutics, however animals may be resistant to developing the arrhythmias required. While the relationships between atrial refractory periods, conduction velocity, surface area, and vulnerability to fibrillation have been established in clinical practice, these parameters are not regularly used to design animal studies of persistent AF (PsAF). Purpose We investigated whether routinely collected baseline parameters could be used to improve experimental efficiency in an ovine model, by predicting the development of PsAF as opposed to arrhythmia resistance. The aims were to: reduce the number of animals used in future studies, and avoid prolonged experiments in animals likely to be resistant to AF. Methods All procedures were conducted with respect to the Animals [Scientific Procedures] Act, UK, 1986; and were approved by the local ethical review board. The ovine model consisted of healthy adult Welsh mountain sheep that underwent implantation of a neurostimulator connected via an endocardial pacing lead to the right atrial appendage. The device was programmed to deliver intermittent 30 second bursts of 50Hz and sheep were monitored over an eight week period for PsAF. Eight variables were collected at time of implant including weight (kg), left atrial diameter (LAD; cm), P wave duration (msec), PR interval (msec), atrial effective refractory period (ERP; msec), atrial conduction velocity (CV; m/s), AF inducibility with 50Hz bursts (secs), and rate threshold of atrial action potential alternans (msec). Analysis of the data was performed using multiple logistic regression and receiver-operator characteristic (ROC) curves. Regression coefficients are presented as natural logarithm of odds ratios (OR) with 95% confidence intervals (CI). Results Seventeen sheep were included in this study. Five (29%) developed PsAF whereas twelve (71%) were resistant (non-sustained or no AF). Univariate analysis found none of the parameters alone could predict PsAF, however ERP (OR -0.05, CI -.01 to 0.01, p = 0.089) and LAD (OR 8.1, CI -1.6 to 17.5, p = 0.095) suggested a combination may be predictive. A multivariate analysis using Fibrillation number (calculated as LAD / [ERP X CV]) was predictive (OR 26.9, CI 1.1 to 52.7; p = 0.04], with an area under ROC curve of 0.85 (p = 0.027). Conclusions Fibrillation number can predict the development of PsAF in healthy sheep. Practically speaking, this means animals with: a larger LAD, shorter ERP and slower CV are more likely to develop PsAF. These findings can be used to optimise the design of future studies, particularly by reducing the number of animals required.